Uncertainty quantification in ion-solid interaction simulations

详细信息    查看全文

• 作者：R. Preuss ; ^{preuss@ipp.mpg.de} ; U. von Toussaint
• 关键词：Polynomial chaos ; Spectral expansion ; Discrete projection ; Uncertainty quantification ; Ion&ndash ; solid interaction ; SDTRIM
• 刊名：Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms
• 出版年：2017
• 出版时间：15 February 2017
• 年：2017
• 卷：393
• 期：Complete
• 页码：26-28
• 全文大小：303 K
• 卷排序：393

文摘

Within the framework of Bayesian uncertainty quantification we propose a non-intrusive reduced-order spectral approach (polynomial chaos expansion) to the simulation of ion–solid interactions. The method not only reduces the number of function evaluations but provides simultaneously a quantitative measure for which combinations of inputs have the most important impact on the result. It is applied to SDTRIM-simulations (Möller et al., 1988) with several uncertain and Gaussian distributed input parameters (i.e. angle, projectile energy, surface binding energy, target composition) and the results are compared to full-grid based approaches and sampling based methods with respect to reliability, efficiency and scalability.